Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A computer-implemented method comprising: defining, by a computing device, an image data identifier, the image data identifier specifying one or more prohibited object types; defining, by the computing device, a data identifier validator, the data identifier validator specifying one or more prohibited object sub-types; receiving, by the computing device, an image; determining, by the computing device, one or more attributes of the image; identifying, by the computing device, one or more objects in the image based on the one or more attributes of the image; determining, by the computing device, an object type of a first object of the one or more objects using the one or more attributes of the image; determining, by the computing device, whether the object type of the first object matches at least one of the one or more prohibited object types; based on determining that the object type of the first object matches at least one of the one or more prohibited object types, determining, by the computing device, whether an object sub-type of the first object matches at least one of the one or more prohibited object sub-types; and based on determining that the object sub-type of the first object matches the prohibited object sub-type, classifying, by the computing device, the first object as prohibited.
Image analysis and content moderation. This invention addresses the need to identify and classify prohibited objects within digital images. A computing device is used to establish criteria for identifying prohibited content. Specifically, an image data identifier is defined, which specifies general categories of objects that are not allowed. Concurrently, a data identifier validator is defined, which further refines these restrictions by specifying particular sub-categories of prohibited objects. The system then receives an image and analyzes its attributes to detect objects present within it. For each identified object, its type is determined based on these attributes. The system checks if the object's type aligns with any of the general prohibited object types. If a match is found, a more granular check is performed to see if the object's sub-type falls within the defined prohibited sub-types. If both the object type and sub-type match the defined prohibitions, the object is definitively classified as prohibited.
2. The computer-implemented method of claim 1 , further comprising: in response to determining that the object type of the first object does not match at least one of the one or more prohibited object types or that the object sub-type of the first object does not match at least one of the one or more prohibited object sub-types, classifying, by the computing device, the first object as allowed.
This invention relates to a computer-implemented method for classifying objects based on their type and sub-type to determine whether they are allowed or prohibited. The method addresses the problem of automatically filtering or permitting objects in a system, such as files, data, or other digital entities, by evaluating their characteristics against predefined criteria. The method involves analyzing a first object to determine its object type and object sub-type. If the object type or sub-type matches any of the prohibited types or sub-types stored in a database, the object is classified as prohibited. Conversely, if the object type and sub-type do not match any prohibited entries, the object is classified as allowed. This classification process ensures that only permissible objects are processed or accessed, enhancing security and compliance in systems where certain object types or sub-types are restricted. The method may also involve comparing the object's attributes against a list of allowed types or sub-types, further refining the classification. The system dynamically checks these attributes in real-time, enabling efficient and accurate filtering. This approach is particularly useful in applications such as cybersecurity, data governance, or content management, where strict control over object types is necessary. The invention improves upon existing systems by providing a more granular and automated way to enforce object classification rules.
3. The computer-implemented method of claim 1 , wherein determining whether the object type of the first object matches at least one of the one or more prohibited object types comprises: determining, by the computing device, a level of confidence that the object type of the first object matches at least one of the one or more prohibited object types; and determining, by the computing device, whether the level of confidence that the object type of the first object matches at least one of the one or more prohibited object types satisfies a threshold level of confidence.
The invention relates to a computer-implemented method for detecting and handling prohibited objects in digital content, such as images or videos. The method addresses the challenge of accurately identifying objects that violate predefined rules, such as inappropriate or restricted items, while minimizing false positives. The system analyzes digital content to detect objects and determines their types using machine learning or pattern recognition techniques. If an object is identified as potentially prohibited, the method calculates a confidence level indicating the likelihood that the object matches any of the prohibited types. This confidence level is then compared against a predefined threshold to decide whether the object should be flagged or blocked. The threshold ensures that only high-confidence matches trigger actions, reducing errors. The method may also involve additional steps, such as generating alerts or modifying the content to remove or obscure the prohibited object. The invention improves content moderation by balancing accuracy and efficiency in automated systems.
4. The computer-implemented method of claim 1 , wherein determining, by the computing device, whether an object sub-type matches at least one of the one or more prohibited object sub-types is further in response to determining that a level of confidence that the object type of the first object matches at least one of the one or more prohibited object types does not satisfy a threshold level of confidence.
This invention relates to computer-implemented methods for detecting and filtering prohibited objects in digital content, such as images or videos, using machine learning models. The problem addressed is the challenge of accurately identifying and blocking objects that violate predefined rules, such as inappropriate or restricted items, while minimizing false positives or negatives. The method involves analyzing a digital content item to detect objects within it. A machine learning model classifies each detected object into an object type and an object sub-type. The system then checks whether the object type or sub-type matches any entries in a predefined list of prohibited object types or sub-types. If a match is found, the object is flagged for further action, such as removal or restriction. A key aspect of the invention is an additional confidence-based filtering step. If the system determines that the confidence level of the object type classification does not meet a predefined threshold, the method proceeds to check whether the object sub-type matches any prohibited sub-types. This ensures that objects are only flagged if there is sufficient certainty in their classification, reducing errors. The method may also involve adjusting the threshold dynamically based on factors like object context or user feedback to improve accuracy over time. The system can be applied in various domains, including content moderation, security screening, or automated compliance checks.
5. The computer-implemented method of claim 1 , further comprising: determining, by the computing device, an object type of a second object of the one or more objects; determining, by the computing device, whether the object type of the second object matches at least one of the one or more prohibited object types; in response to determining that the object type of the second object matches at least one of the one or more prohibited object types, determining, by the computing device, whether an object sub-type of the second object matches at least one of the one or more prohibited object sub-types; and in response to determining that the object sub-type of the second object matches the prohibited object sub-type, classifying, by the computing device, the second object as prohibited.
This invention relates to computer-implemented methods for identifying and classifying prohibited objects within a digital environment. The method addresses the challenge of accurately detecting and categorizing objects that violate predefined rules or policies, such as restricted items in an image or data set. The system processes one or more objects, analyzing their characteristics to determine compliance with allowable criteria. The method involves determining the object type of a second object among the processed objects. It then checks whether this object type matches any of the predefined prohibited object types. If a match is found, the system further examines the object's sub-type to verify if it aligns with any prohibited sub-types. If both the type and sub-type are prohibited, the object is classified as such. This hierarchical classification ensures precise identification of restricted items, reducing false positives and improving enforcement of digital content policies. The approach is particularly useful in applications like content moderation, security screening, or compliance monitoring, where accurate object detection is critical.
6. The computer-implemented method of claim 5 , further comprising: determining, by the computing device, that the image contains sensitive information based on the object type of the first object and the object type of the second object.
This invention relates to computer-implemented methods for detecting sensitive information in images. The problem addressed is the need to automatically identify and flag images containing sensitive data, such as personal or confidential information, to prevent unauthorized access or disclosure. The method involves analyzing an image to detect objects within it, where each object is classified into an object type. The system determines whether the image contains sensitive information by evaluating the object types of at least two detected objects. For example, if the image contains a face (object type: "person") and a document (object type: "ID card"), the system may infer that the image includes sensitive personal information. The method may also involve additional steps, such as comparing the detected objects to predefined categories of sensitive information or applying machine learning models to assess the context of the objects. The goal is to provide an automated way to identify and handle images that may pose privacy or security risks.
7. The computer-implemented method of claim 6 , wherein it is determined that the image contains sensitive information responsive to the object type of the first object being a first particular object type and the object type of the second object being a second, different particular object type.
This invention relates to computer-implemented methods for detecting sensitive information in images. The problem addressed is the need to automatically identify and flag images containing sensitive data, such as personally identifiable information (PII), before they are shared or processed further. Existing solutions may rely on manual review or basic pattern recognition, which can be inefficient or inaccurate. The method involves analyzing an image to detect multiple objects, each associated with an object type. The system determines whether the image contains sensitive information based on the presence of at least two distinct object types. For example, if the first object is identified as a passport (a first particular object type) and the second object is identified as a face (a second, different particular object type), the system concludes that the image likely contains sensitive information. The method may also involve comparing the detected objects to a predefined list of sensitive object types to confirm the presence of sensitive data. This approach improves accuracy by leveraging contextual relationships between objects rather than relying solely on individual object detection. The system can then trigger actions such as blocking the image, redacting sensitive portions, or alerting a user. This solution enhances privacy and compliance in digital workflows by automating the detection of sensitive content in images.
8. The computer-implemented method of claim 5 , further comprising: determining, by the computing device, that the image contains sensitive information based on the object sub-type of the first object and the object sub-type of the second object.
This invention relates to computer-implemented methods for detecting sensitive information in images. The problem addressed is the need to automatically identify and flag images containing sensitive data, such as personal or confidential information, to prevent unauthorized disclosure. The method involves analyzing an image to detect objects within it, where each detected object is classified into an object type and an object sub-type. The object type represents a broad category (e.g., text, face, document), while the object sub-type provides a more specific classification (e.g., passport, ID card, medical record). The system then evaluates the combination of object sub-types in the image to determine whether the image contains sensitive information. For example, if the image contains a face sub-type classified as a "passport photo" and a text sub-type classified as "passport number," the system would flag the image as containing sensitive information. The method may also include additional steps, such as extracting the sensitive information from the image, redacting it, or alerting a user. The system can be trained using machine learning models to improve accuracy in detecting and classifying objects and their sub-types. This approach ensures that sensitive data is identified and handled appropriately, reducing the risk of data breaches.
9. The computer-implemented method of claim 1 , further comprising: automatically blocking the image from being electronically transferred in response to the first object being classified as prohibited.
This invention relates to automated image analysis for content moderation, specifically preventing the transfer of images containing prohibited objects. The system uses computer vision techniques to analyze digital images and identify objects within them. When an image is processed, the system classifies objects detected in the image. If any object is classified as prohibited, the system automatically blocks the image from being electronically transferred. This prevents unauthorized or inappropriate content from being shared. The system may also include additional steps such as generating alerts, logging violations, or flagging the image for review. The technology is designed for applications in social media, messaging platforms, or other digital communication systems where content moderation is required. The automated blocking mechanism ensures compliance with legal or platform-specific content policies without manual intervention. The system may integrate with existing image processing pipelines or operate as a standalone moderation tool. The invention addresses the challenge of efficiently detecting and preventing the distribution of prohibited content in digital communications.
10. The computer-implemented method of claim 1 , further comprising: automatically transmitting an electronic message to an administrator in response to the first object being classified as prohibited.
A computer-implemented method addresses the problem of detecting and managing prohibited content in digital systems. The method involves classifying objects, such as files or data, to determine if they are prohibited based on predefined criteria. When an object is classified as prohibited, the system automatically transmits an electronic message to an administrator. This notification allows the administrator to take appropriate action, such as removing the prohibited content or investigating the source. The classification process may involve analyzing the object's attributes, metadata, or content to identify violations of policies or regulations. The method ensures timely detection and response to prohibited content, enhancing security and compliance in digital environments. The system may also include additional steps, such as logging the incident, quarantining the object, or generating reports for auditing purposes. The automated notification reduces manual oversight, improving efficiency in managing prohibited content.
11. The computer-implemented method of claim 1 , wherein the one or more attributes of the image comprise dimensions of the first object.
The invention relates to computer-implemented methods for analyzing images to determine attributes of objects within them. The method addresses the challenge of accurately extracting specific characteristics, such as dimensions, from objects in digital images, which is crucial for applications like object recognition, measurement, and quality control in manufacturing or automation. The method involves processing an image containing at least one object to identify and measure its dimensions. This includes detecting the object within the image, isolating it from other elements, and calculating its physical dimensions based on pixel data. The process may involve techniques such as edge detection, contour analysis, or machine learning models trained to recognize and measure object boundaries. The extracted dimensions can then be used for further analysis, such as comparing the object to predefined specifications or integrating the data into a larger system for decision-making. The method ensures precise and automated measurement of object dimensions, reducing human error and improving efficiency in tasks that require dimensional accuracy. This is particularly useful in industries where consistent and reliable measurements are critical, such as robotics, industrial automation, and quality assurance. The approach leverages computational techniques to enhance the accuracy and speed of dimensional analysis in digital images.
12. The computer-implemented method of claim 1 , wherein the one or more attributes of the image comprise MICR (Magnetic Ink Character Recognition) characters.
This invention relates to computer-implemented methods for processing images, specifically focusing on the extraction and analysis of MICR (Magnetic Ink Character Recognition) characters. MICR technology is widely used in banking and financial applications to identify and process checks, ensuring accurate and secure transactions. The challenge addressed by this invention is the reliable detection and interpretation of MICR characters in digital images, which can be affected by variations in print quality, background noise, or image distortion. The method involves analyzing an image to identify and extract MICR characters, which are printed using a special magnetic ink that enhances readability by machines. The system processes the image to isolate these characters, applying techniques to filter out non-MICR elements and enhance the clarity of the detected characters. This ensures that the extracted MICR data is accurate and can be used for further processing, such as verifying bank account information or routing numbers. The invention may also include additional steps to preprocess the image, such as noise reduction, contrast adjustment, or skew correction, to improve the accuracy of MICR character recognition. The extracted MICR data is then validated against known standards to confirm its correctness before being used in financial transactions or other applications. This method ensures that MICR characters are reliably detected and interpreted, reducing errors in automated check processing systems.
13. The computer-implemented method of claim 1 , wherein the one or more attributes of the image comprise an image of a face.
The invention relates to computer-implemented methods for analyzing images, specifically focusing on facial recognition and attribute extraction. The method involves processing an image to identify and extract one or more attributes, with a particular emphasis on facial features. The system captures an image, such as a photograph or video frame, and applies image processing techniques to detect and analyze facial characteristics. These attributes may include facial landmarks, expressions, or other distinguishing features. The method may also involve comparing the extracted attributes against a database or reference set to identify matches or similarities. This technology is useful in applications like security systems, biometric authentication, and personalized user experiences. The method ensures accurate and efficient facial recognition by leveraging advanced image processing algorithms, which enhance the reliability of attribute extraction in various lighting and environmental conditions. The system may also integrate machine learning models to improve recognition accuracy over time. This approach addresses challenges in facial recognition, such as variations in lighting, pose, and occlusions, by employing robust feature extraction techniques. The method is designed to work with different types of images, including still images and video streams, making it versatile for real-world applications.
14. The computer-implemented method of claim 13 , wherein the one or more attributes of the image comprise the image of the face at a defined location on the first object.
This invention relates to computer vision and object recognition, specifically improving the accuracy of detecting and analyzing facial images on physical objects. The problem addressed is the difficulty in reliably identifying and processing facial images when they appear on objects, where variations in lighting, angle, and background can obscure or distort the face. The solution involves a method that extracts and analyzes specific attributes of an image, particularly focusing on the position of a face relative to a defined location on an object. By precisely locating the face within a structured reference frame, the system enhances recognition accuracy and reduces errors caused by environmental factors. The method may also include preprocessing steps to enhance image quality, such as noise reduction or contrast adjustment, before analyzing the facial attributes. The invention is particularly useful in applications like security systems, augmented reality, and automated surveillance, where accurate face detection on objects is critical. The approach ensures that the face is consistently identified at a predetermined position, improving reliability in dynamic environments.
15. The computer-implemented method of claim 14 , wherein the object type of the first object comprises a photo identification card.
A system and method for processing and analyzing photo identification cards in a digital environment. The technology addresses the challenge of securely and accurately verifying and extracting information from photo identification cards, such as driver's licenses, passports, or employee badges, in automated systems. The method involves capturing an image of the photo identification card using a digital imaging device, such as a camera or scanner. The system then processes the image to detect and extract relevant data, including text, barcodes, or embedded digital signatures, to verify the card's authenticity and retrieve personal information. The system may also compare the extracted data against a database or perform facial recognition to confirm the identity of the cardholder. Additionally, the method may include encrypting the extracted data for secure transmission or storage. The system is designed to handle various formats and standards of photo identification cards, ensuring compatibility across different jurisdictions and issuers. The technology is particularly useful in applications requiring identity verification, such as access control, financial transactions, or digital onboarding processes.
16. The computer-implemented method of claim 1 , wherein the one or more prohibited object types comprise a driver license and the one or more prohibited object sub-types comprise a particular driver license issuing jurisdiction.
This invention relates to computer-implemented methods for detecting and restricting the capture or processing of sensitive objects, such as identification documents, in digital images or video streams. The method addresses the problem of unauthorized or unintended capture of personal identification documents, such as driver licenses, which may contain sensitive information. The system identifies prohibited object types, such as driver licenses, and further classifies them into specific sub-types based on issuing jurisdictions to enforce stricter controls. The method involves analyzing digital media to detect the presence of these prohibited objects, then applying restrictions to prevent their capture, storage, or transmission. The system may also generate alerts or notifications when such objects are detected. The invention is particularly useful in applications where privacy and data protection are critical, such as in surveillance systems, document scanning, or social media platforms. By distinguishing between different issuing jurisdictions, the method allows for tailored enforcement policies, ensuring compliance with regional regulations. The system may integrate with existing image processing or computer vision tools to enhance detection accuracy and efficiency. The overall goal is to mitigate privacy risks by preventing the unauthorized dissemination of sensitive identification documents.
17. A computer readable medium storing instructions that, when executed by a computing device having one or more processors, causes the one or more processors to perform operations comprising: defining, by the computing device, an image data identifier, the image data identifier specifying one or more prohibited object types; defining, by the computing device, a data identifier validator, the data identifier validator specifying one or more prohibited object sub-types; receiving, by the computing device, an image; determining, by the computing device, one or more attributes of the image; identifying, by the computing device, one or more objects in the image based on attributes of the image; determining, by the computing device, an object type of a first object of the one or more objects using the one or more attributes of the image; determining, by the computing device, whether the object type of the first object matches at least one of the one or more prohibited object types; determining, by the computing device, whether an object sub-type of the first object matches at least one of the one or more prohibited object sub-types based on a determination that the object type of the first object matches the at least one of the one or more prohibited object types; and in response to determining that the object sub-type of the first object matches the at least one of the one or more prohibited object sub-types, classifying, by the computing device, the first object as prohibited.
This invention relates to image analysis systems for detecting and classifying prohibited objects within images. The system addresses the challenge of accurately identifying and categorizing objects that violate predefined rules, such as restricted items in surveillance or content moderation applications. The system uses a computing device to process images by first defining an image data identifier that specifies prohibited object types, such as weapons or illegal substances. Additionally, a data identifier validator is defined to specify prohibited object sub-types, such as specific weapon models or drug variants. Upon receiving an image, the system analyzes its attributes to detect objects and determine their types. If an object matches a prohibited type, the system further checks if its sub-type is also prohibited. If both conditions are met, the object is classified as prohibited. This hierarchical approach ensures precise identification of restricted items, improving accuracy in automated content filtering and compliance enforcement. The system leverages machine learning or pattern recognition techniques to analyze image attributes and classify objects, enabling real-time or batch processing of visual data.
18. The computer readable medium of claim 17 , wherein the operations further comprise: in response to determining that the object type of the first object does not match at least one of the one or more prohibited object types or that the object sub-type of the first object does not match at least one of the one or more prohibited object sub-types, classifying, by the computing device, the first object as allowed.
This invention relates to a system for classifying objects in a computing environment, particularly for determining whether an object should be allowed or prohibited based on its type and sub-type. The system addresses the problem of unauthorized or malicious objects accessing or interacting with computing resources by implementing a filtering mechanism that checks object attributes against predefined rules. The system operates by analyzing a first object to determine its object type and object sub-type. The object type and sub-type are then compared against a list of prohibited object types and sub-types stored in a database. If the object's type or sub-type matches any entry in the prohibited list, the object is classified as prohibited. Conversely, if the object's type and sub-type do not match any prohibited entries, the object is classified as allowed. This classification process ensures that only authorized objects are permitted to proceed, enhancing security in the computing environment. The system may also include additional operations, such as retrieving the prohibited object types and sub-types from a database and performing the classification in response to a request to access a resource. The classification decision is based on whether the object's attributes match any prohibited entries, providing a flexible and scalable approach to object filtering. This method improves security by preventing unauthorized or malicious objects from accessing sensitive resources.
19. A computer system comprising: an identifier definition module programmed to define an image data identifier and a data identifier validator, the image data identifier specifying one or more prohibited object types, the data identifier validator specifying one or more prohibited object sub-types, at least one processor configured to execute the identifier definition module; a receiving module programmed to receive an image, the at least one processor configured to execute the receiving module; an object identification module programmed to determine one or more attributes of the image identify one or more objects in the image based on the one or more attributes of the image, the at least one processor configured to execute the object identification module, the object identification module communicatively coupled to the receiving module; an object analysis module programmed to determine an object type of a first object of the one or more objects using the one or more attributes of the image, the at least one processor configured to execute the object analysis module, the object analysis module communicatively coupled to the object identification module; and a classification module programmed to determine whether the object type of the first object matches at least one of the one or more prohibited object types, in response to determining that the object type of the first object matches at least one of the one or more prohibited object types, determine whether an object sub-type of the first object matches at least one of the one or more prohibited object sub-types, the object sub-type being determined based on the object type, and, in response to determining that the object sub-type of the first object matches the at least one of the one or more prohibited object sub-types, classify the first object as prohibited, the at least one processor configured to execute the classification module, the classification module communicatively coupled to the object analysis module and the identifier definition module.
A computer system analyzes images to detect and classify prohibited objects. The system includes modules for defining prohibited object types and sub-types, receiving images, identifying objects within images, analyzing object attributes, and classifying objects based on predefined rules. The identifier definition module specifies prohibited object types and sub-types, while the receiving module captures input images. The object identification module processes image attributes to detect objects, and the object analysis module determines the type of each detected object. The classification module checks whether an object's type matches any prohibited type and, if so, further checks if its sub-type is also prohibited. If both conditions are met, the object is classified as prohibited. This system enables hierarchical filtering of objects in images, allowing for granular control over what is considered prohibited based on both broad categories and specific sub-categories. The hierarchical approach improves accuracy by reducing false positives and ensuring only truly prohibited objects are flagged. The system is useful for applications requiring strict content moderation, such as social media platforms, security systems, or automated surveillance.
20. The computer system of claim 19 , wherein the classification module is further programmed to, in response to determining that the object type of the first object does not match at least one of the one or more prohibited object types or that the object sub-type of the first object does not match at least one of the one or more prohibited object sub-types, classify the first object as allowed.
This invention relates to a computer system for classifying objects based on their type and sub-type to determine whether they are allowed or prohibited. The system addresses the problem of automatically identifying and filtering objects that violate predefined rules, such as security policies or content restrictions, by analyzing their attributes. The system includes a classification module that processes objects by comparing their type and sub-type against a list of prohibited types and sub-types. If the object's type or sub-type matches any prohibited entry, the object is classified as prohibited. Conversely, if neither the type nor the sub-type matches any prohibited entry, the object is classified as allowed. The system may also include a detection module that identifies objects within a data stream or storage system and extracts their type and sub-type information for classification. The classification module may further apply additional rules or thresholds to refine the classification process. This invention is useful in applications such as cybersecurity, content filtering, and compliance monitoring, where automated object classification is required to enforce policies or detect violations.
Unknown
November 26, 2019
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.